I've decided to start a new regular segment on Accrued Interest doing technical analysis of the 10-year Treasury. I'm planning on doing this about weekly, although I won't swear it will always be on the same day.
This will be something of a departure for this blog, since classically I've rarely talked about trades I'm actually doing and even more rarely talked about short term trades I'm actually doing. But here is why I'm doing it. I think that going forward, there will be more money to be made trading bonds with a short-term view. I believe volatility will be permanently higher than in the recent past, and the ability to discern short-term movements will be the key to making money. I'd like to think opening up a discussion will help readers make more profitable trades.
The other reason is purely selfish. It is my long-term goal to start my own hedge fund. It wouldn't really be for the money, more because I think that's how I'm wired. If I could be running my own hedge fund but my income was only 75% of what it is now, I'd make that trade in a second. Anyway, although this eventual move may be a few years away, I've already built many of the models I plan to use as part of my trading strategy. Treasury technicals isn't a model per se but it would be part of my trading strategy. The models are proprietary, but technicals aren't. And besides, by sharing my strategies, I hope to get feedback from readers to perhaps improve my results.
So let's get to it. You may remember last week I showed a chart suggesting that yields on 10s ought to fall. Didn't happen, so let's review what actually did. Below is a similar chart to the one from last week updated to now.
Note this is a yield chart, so downward moves are bullish. The yellow line is a 200D SMA, the red and green were trend lines I drew. At the time, the red and green seemed to suggest to me a pattern of lower lows and higher highs. But now it looks like a wedge. It bounced off the SMA, broke above the trend line and would seem to be headed higher in yield, lower in price.
The above is a 30D intra-day price chart for the current 10-year. You can see the gap downward on 10/9, 10/12 was a holiday, 10/13 we filled most of the gap, and then started moving lower again.
Looking at a plan vanilla MACD chart, short-term momentum is clearly bearish.
So the trade looks like a short. What's the entry? Seems like there is resistance at 3.48%. This bar chart (with Fibs) shows there has been a lot of work done right around the current price level (this is a price chart, I know, price/yield gets confusing. You'll get used to it.) I've circled the price area of approximately 101-3 to 101-16, where we see lots of action in early-mid September. That corresponds to about a 3.45-3.48% yield. Right where we are now.
Note that bond guys didn't seem to give a shit about the 50% re-trace from the highs. I know a lot of bond guys follow Fibonacci re-traces, but I honestly haven't been able to make it work for me all too often. Don't be fooled by the fact that 3.48% is very near the 38.2% line, because when all those trades were happening, it wasn't the 38.2% line! Only once we hit the high on 10/2 did that become the number.
3.48% is backed up by this intra-day yield chart. The yellow line is 3.48%.
Friday, October 16, 2009
Don't get technical with me! 10/16/09
After doing all this, my play is to leg into a short in three parts. First I've put on some right now. Second I may get the chance to put on more at about 101-24, or 3.41%,. Again, looking at the intra-day, that looks like an area of high-volume that we could revisit on a bounce off the 3.48% resistance. Then I'd add a third chunk once we breach that resistance in a meaningful way, maybe 3.50%.
Subscribe to:
Post Comments (Atom)
19 comments:
Remember the wisdom I learned as a young quant, when markets trade technically, one man's technical buy has to be another man's technical sell...
The hard nosed economics that insists technicals are invalid has been demonstrated to be suboptimal. However, realistically fundamentals will always overwhelm technicals.
I like your analysis. The only caveat is that on a weekly chart of the 10-year yield the stochastic indicator is oversold and turning up leaving open the possibility of a sharp rally (that would be a sell-off in the bond). I don't trade bonds so I don't know if you guys use stops but I would get out at least partially should the yield break decisively above 3.60.
I am curious as to your thoughts on new regulatory requirements for banks to have higher capital ratios.
Under Basel II, banks are supposed to have higher ratios of "safe" assets (aka treasuries). But US banks have fought tooth and nail to maintain their hedges in the OTC derivatives market by arguing this is a more "competitive" advantage instead of wasting money on averaging down on a declining asset like bonds.
There is also talk of increased deal flow in hybrid bonds. Wouldn't such activity satisfy liquidity buffers too?
The American Bar Association is having their fall meeting in Washington DC w/a big focus on derivatives and banking ratios. It will be interesting to see what the attorneys have to say about future projections. I will try to be there either through teleconference or in person. Worst case scenario, I can always get the materials from the committees that upload it online (slight delay there).
In the meantime, what are your thoughts on Basel II, implementing higher capital ratios, and hybrid bonds? Will this lead to greater demand for treasuries (at least on the short-mid end of the curve)?
Obviously in light of what's happened, its hard to argue against higher capital ratios. But here's what concerns me. A very large percentage of Citi's problems were off-balance sheet. A lot of the bad lending went into securitized vehicles. I'm afraid whatever rules you devise, banks will figure a way around it. I'd rather live in a world of less capital-related regulations and more regs focused on the too-big-to-fail problem. If a bank can take risks and fail without contagion, then we have a better system. If we set up a bunch of new rules, we just get new cheaters.
AI, the Treasury put this up for notice and comment:
http://www.occ.treas.gov/ftp/
release/2009-101a.pdf
Determining whether a specific company is required to consolidate a VIE under FAS 167 depends on a qualitative analysis of whether that company has a “controlling financial interest” in the VIE. The analysis focuses on the company’s power over and interest in the VIE, rather than on quantitative equity ownership thresholds. A company has a controlling financial interest in a VIE if it has (1) the power to direct matters that most significantly impact the activities of the VIE, including, but not limited to, activities that impact the VIE’s economic performance (for example, servicing activities); and (2) either the obligation to absorb losses of the VIE thatpotentially could be significant to the VIE, or the right to receive benefits from the VIE that potentially could be significant to the VIE, or both.
I wouldn’t trust any of these market timers: I’m not a technical trader:
Robert Pretcher, Joseph Granville, Richard Dennis, Welles Wilder, Paul Tudor Jones, William O'Neil, Marty Schwartz, Justin Mamis, Edwards and Magee, William Bretz.
Fundamentals rule the markets, not technical’s. Traders don’t know the fundamentals. So you end up having Joseph Granville proclaiming “technicals precede fundamentals”.
Don't bet on it.
TRADING USING FUNDAMENTALS:
=================================
flow5 (2/26/07; 14:34:35MT - usagold.com msg#: 152672)
Suckers Rally
If gold doesn't fall, then there's a new paradigm. The drop in member commercial bank adjusted "free" legal reserves is unprecedented
==================================
The DOW fell 3.3% & 416 points (the 7th largest poinnt drop ever)
================================
Some people think Feb 27, 2007
started across the ocean:
"On Feb. 28, Bernanke told the House Budget Committee he could see no single factor that caused the market's pullback a day earlier".
In fact, it was home grown. Feb 27coincided with the sharpest decline in:
=================================
"Today, with bank reserves largely driven by bank payments (debits), your views on bank debits and legal reserves sound right!" - Senior Vice President, St. Louis FED
Contrary to economic theory, the monetary lags for real-growth & inflation are fixed in length.
However the lag for nominal gdp varies widely. I.e., a lot of Ph.Ds only use one lag, the wrong one (which can only be nominal gdp?)
=================================
If the results are incongruous, it's not the theory, its the FED's data. It's mal-adjusted, it's it's non-conforming, it's unobtainable, etc. Try putting together a "time series" with the FED's statistics, i.e., the data is revised, reconstructed, then overlaid.
Contrary to Milton Friedman, the "monetary base" is not a base for the expansion of money and credit, legal reserves are. Currency has no expansion coefficient.
Banks don't loan out excess reserves ($860b), they are earning assets.
Any institution whose liabilities can be transferred on demand, without notice, and without income penalty, by data networks, checks, or similar types of negotiable credit instruments, and whose deposits are regarded by the public as money, can create new money, provided that the institution is not encountering a negative cash flow.
From a systems viewpoint, commercial banks as contrasted to financial intermediaries: never loan out, and can’t loan out, existing deposits (saved or otherwise) including existing transaction deposits (TRs), or time deposits (TDs) or the owner’s equity, or any liability item.
When CBs grant loans to, or purchase securities from, the non-bank public, they acquire title to earning assets by initially, the creation of an equal volume of new money- (Transaction deposits -TRs) -- somewhere in the banking system.
I.e., commercial bank deposits are the result of lending, not the other way around.
The non-bank public includes every institution, the U.S. Treasury, the U.S. Government, State, and other Governmental Jurisdictions, and every person, except the commercial and the Reserve banks.
The basic expansion coefficient (my definition), for the banking system as a whole (the correct source-base), is obtained by dividing commercial bank credit (or the Assets and Liabilities of Commercial Banks in the United States H.8), by primaryliy, the sum of the member bank’s:
(1) required reserves, plus (2) contractual clearing balances (reductions in required reserve balances have predominantly been accompanied by “offsetting increases” in the member bank’s contractual (required), clearing balances plus (3) supplemental reserves (daylight overdraft credit).
However figures for daylight credit are only available with a large lag - figures.
Another example: Black monday, Oct 19, 1987. It wasn't program trading, it wasn't a black swan (Robert Prechter nailed it). The FED ran an excessive and unnecessary "tight" monetary policy. Interest rates were already very constrictive.
Historically, Black Monday was the largest contraction in legal reserves ever recorded (but not by the FED, by the monetary lag for real-growth). Monetary lags are always the same length.
Will you be legging into your shorts via actual bonds or etf's?
Why not CBOT bond futures? I haven't had one losing trade since july 1979 (as long as there were contracts. Just luck.
Note also: So many people followed Granville that they anticipated his calls. That ruined his record for the next 19 years. However he is hot as a firecracker since 2000, i.e, he doesn't call turns, he calls the day the market turns.
thanks for this new segment and good look with that (and your HF plans).
but a question to start with:
since the treasury market is one of the most battled ones - what are your past experiences?
does 101 technical analysis really work here?
or is more like in forex or the major stock indexes - a lot of faking and you have to be an artist to make profits from technical trading?
Here are a few thoughts off the top of my head. These are based on my own experience: I spent the best part of a decade trading Treasuries for a hedge fund, mostly quantitative relative value but I did my share of macro and directional stuff as well.
1. I like that you use very simple indicators (trendlines, wedges, support and resistance). In my experience more complex technical indicators don’t work very well in the Treasury market. Besides with more complex indicators you have to constantly guard against over-fitting the data.
2. That said, I’ve found the Japanese candlestick or ichimoku tool (Bloomberg GOC) to be useful especially in gauging the *strength* (as opposed to location) of a given support or resistance level.
3. I think you make a convincing case that 3.48 is important support and a break thereof would be bearish. I don’t have any quibbles with your analysis per se.
4. But I do have a quibble with your proposed trading strategy. If I understand your post correctly, you established a short before the market broke through its support. I think this is a mistake. I always prefer to go short after the break happens. Sure I may give up a few ticks initially but I don’t mind since I’m typically looking to make a few tens of ticks. On the other hand if I go short too early and the support holds then I’m just going to be whipsawed.
5. In fact, as long as we’re above the support line, my first instinct would be to buy not sell. I view a support line as valid until it’s broken. After it’s broken, the 3.48% line is now resistance, so that’s when I go short.
6. By the same token. I think it’s a mistake to add if the market goes against you (rallies to 101-24 or 3.41%). To me, technical trading is all about finding the path of least resistance for the market. If 10s bounce off 3.48% instead of breaking it, that means the path of least resistance is higher (ie lower yields). Don’t let your preconceived notions (that the market is poised for a selloff) prevent you from taking the right action – in this case, buying.
Okay, so blogger won’t let me post my entire comment in one go (character limit), so the rest of this comment follows separately.
(Comment continued). That’s it for the specific critiques of this particular trade. Here are some more general philosophical thoughts on trading directionally:
7. Keep a trading diary. This is vital. Every time you do a trade, write down what you did and why you did it. And it’s important to do this as soon as possible after executing the trade (don’t wait till the end of the day or, heaven forbid, the next morning). This will help your learning and also guard against self-deception.
8. Have a target, either time or profit or both, and a stop-loss. Most of my technical trades lasted 1 week or less (often just 1-3 days). Typically I’d take profits after a 1 point move in 10s (say 10-12 bps). Sometimes I’d take profits earlier (if the trade no longer felt right), sometimes later (if things were rolling nicely). I usually set my stops around 5 ticks (2bps) away from my entry. Of course these numbers may no longer be very useful – the bond market has certainly changed in recent times – so you have to find what works for you.
9. Don’t over-step your time horizon or over-rule your stop-loss no matter what the reason. An especially dangerous trap is holding on to a losing technical trade because “the fundamentals are on your side in the long run”.
10. Forget about shadow-trading, it’s useless. Successful technical trading is all about psychology (your own, and the markets). You’ll never be able to replicate the experience with an imaginary portfolio. Instead, I suggest beginning with trades that are big enough to matter on a day-to-day basis, but small enough that your core (non-technical) portfolio won’t be in danger. For instance, if you have $50mm in capital, start with 5000 of dv01 per trade; that way even 100bps of loss over the entire year will only be 1% in return terms.
11. Once you’re confident trading chunks of 5k, double your size. Then when you’re comfortable with that, double it again. Keep doubling until your size is commensurate with your PL expectation.
12. Trust your instinct. But be aware that you have to watch the market for every tick of every day for years and years before you develop a reliable instinct. Instinct is just distilled experience; a ‘muscle memory’ for the tape. Too many people invoke ‘gut feeling’ without putting in the hard yards required to gain that instinct.
13. To give you an example: I spent 5 years trading Treasuries full-time before trying my first outright trade. My first few outrights all lost money, it was back to the drawing board. Then gradually I got better. It took another 2 years before I was getting things consistently right. Don’t rush this process. (And like I said, it’s vital to trade real money, otherwise the learning simply will not take).
13. Don’t be afraid not to trade. Don’t be afraid to admit you’re wrong. Don’t be afraid to get out of all your positions.
14. Read the masters. I re-read “Reminiscences of a Stock Operator” every 2-3 years, religiously. Every time I re-read it I’m amazed at how much I learn.
Okay, well, this comment turned out to be a lot longer than I planned, but I hope you find it useful. Feel free to email me, metafinblog @ gmail, if you want to follow up any of these points.
BTW this is my first time commenting here; just wanted to say I do enjoy reading your blog very much -- thanks very much for writing it!
Cheers!
Meta:
Thanks for the comments. VERY FAIR about not waiting for the breech of 3.48%. I agree with you, and I'm adding some other thoughts into the trade. I'd love it if you'd keep commenting as I keep posting. I'll never get insulted if you critique me using logic!
About the Tsy market being very efficient... there is a flip side and that is that there are lots of non-economic buyers. Hell, all the CB's don't care about level.
I am using ETFs and real bonds.
AI,
What do you think about the TED Spread these days. 20bps? Is there any risk left?
hi ttdg
i've done some probability curves on the UST 10Y.
will try to post the range once I get the chance to update the latest data.
Also, you can use TD indicators if you want to trade short term. It's a good technical tool to use
Cheers
fred
Post a Comment